September 2013 Department of Agriculture and Food

Report card on sustainable natural resource use in agriculture Status and trend in the agricultural areas of the south-west of

Supporting your success Dryland salinity i i Disclaimer The Chief Executive Officer of the Department of Agriculture and Food and the State of Western Australia accept no liability whatsoever by reason of negligence or otherwise from use or release of this information or any part of it. Copies of this document may be available in alternative formats upon request. 3 Baron-Hay Court South WA 6151 Tel: (08) 9368 3333 Email: [email protected] Website: www.agric.wa.gov.au Copyright © Western Australian Agriculture Authority, 2013

Cover photo: Harvesting the wheat crop at Mingenew. 2.7 Dryland salinity

Key messages

Condition and trend More than one million hectares of agricultural land in the south-west of WA is severely salt-affected. Dryland salinity has expanded throughout most of the south-west of WA since 1998, especially following episodic rainfall events, such as occurred in 1999, 2000, 2005, 2006 and 2007. In areas cleared and developed for agriculture after 1960, most watertables continue to rise, despite a decline in annual rainfall. As these areas approach a ‘new’ hydrological equilibrium, climate impacts will become the dominant controller of groundwater level trends and the extent of dryland salinity. Hydrological equilibrium and the potential extent of dryland salinity may take many decades to develop, especially in drier areas. Management implications The implications of dryland salinity to the agricultural industry are widespread and include reductions in crop yield, area of arable land and land capability. The opportunity cost of lost agricultural production is at least $344 million per year. Salinity also physically and financially impacts rural infrastructure, public and private and biodiversity, with costs exceeding those to agriculture. Dryland salinity is a major cause of land degradation and remains a potential threat to 2.8–4.5 million hectares of highly productive, low-lying or valley , across the south-west of WA. Management to contain or adapt to salinity is technically feasible using plant-based and engineering options, though recovery is economically viable in only a few areas.

110 111 1 3 2 Dryland salinity 4 Very Low Low 5 Moderate High

6 7 9 8 Merredin

Perth

13 15 14 11 10 Bunbury 12 25 23 24 16 22 21 17 20 Esperance 18 19 0 100 200300 Albany Kilometres

2.7.1 Resource risk summary for risk of expansion of dryland salinity within hydrozones.

110 111 Table 2.7.1 Resource status and trends summary for dryland salinity

Hydrozone Summary Risk and groundwater trends Confidence Very Very In In High Mod Low high low condition trend 1 Kalbarri Sandplain Extent of salinity very minor. Salinity could develop in the medium term. Variable trends in groundwater levels. 2 Northampton Block Extent of salinity minor and unlikely to expand. Mostly rising groundwater trends as levels recover from drought. 3 East Binnu Sandplain Extent of salinity minor but will almost certainly develop in the medium term because groundwater trends are mostly rising. 4 Irwin Terrace Extent of salinity moderate, no imminent threat of expansion. Variable trends in groundwater levels. 5 Arrowsmith Extent of salinity minor. Salinity could develop in medium term. Variable trends in groundwater levels. 6 Dandaragan Plateau Extent of salinity minor but actively spreading with mostly rising trends in groundwater levels. Large area of high quality agricultural land at risk. 7 Northern Zone of Salinity is extensive and continues to expand, but more slowly than prior to 2000. Ancient Variable trends in groundwater levels. 8 Northern Zone of Extent of salinity moderate and likely to expand but at slower rates than prior to Rejuvenated Drainage 2000. Variable trends in groundwater levels. 9 Southern Cross Extent of salinity moderate with a low risk of expansion because trends in groundwater levels are mostly falling. 10 South-eastern Zone of Salinity is extensive and will continue to expand, albeit at a slower rate. Variable Ancient Drainage trends in groundwater levels, but water levels in upland bores are still rising. 11 South-western Zone Salinity is extensive and likely to expand because groundwater levels are mostly of Ancient Drainage rising. 12 Southern Zone of Salinity is extensive. Groundwater levels are mostly stable. The risk is moderating Rejuvenated Drainage as catchments approach equilibrium. 13 Eastern Darling Extent of salinity moderate, area of future expansion minor. Trends in Range groundwater levels are mostly stable. 14 Western Darling Extent of salinity minor, with low risk of expansion because groundwater levels Range are mostly falling. Zone mostly forested. 15 Coastal Plain Extent of salinity very minor; future development possible in the north, but with low impact. Mostly stable trends in groundwater levels.

112 113 Table 2.7.1 Resource status and trends summary for dryland salinity (cont.)

Hydrozone Summary Risk and groundwater trends Confidence Very Very In In High Mod Low high low condition trend 16 Donnybrook–Leeuwin Extent of salinity very minor and future development is unlikely because the zone is at hydrological equilibrium. 17 Scott Coastal Plain Extent of salinity very minor. Future development is unlikely because the zone is at hydrological equilibrium. 18 Warren–Denmark Extent of salinity minor with variable trends in groundwater levels. Salinity risk is moderate and extent will depend on future land use. 19 Albany Sandplain Extent of salinity minor. Groundwater levels are currently deep with variable trends. 20 Stirling Range Salinity is extensive and groundwater trends are variable and equilibrium will be reached in the short term. 21 Pallinup Extent of salinity moderate; expansion is likely but extent restricted. Variable trends in groundwater levels. 22 Jerramungup Plain Extent of salinity moderate, expanding slowly and likely to continue. Mostly rising trends in groundwater levels. 23 Ravensthorpe Extent of salinity minor; expansion is likely but extent restricted. Mostly rising trends in groundwater levels. 24 Esperance Sandplain Extent of salinity moderate but will almost certainly continue to expand because groundwater levels are mostly rising. 25 Salmon Gums Mallee Extent of salinity moderate; expansion is possible because groundwater levels are mostly rising.

Salinity risk grades Salinity risk grades matrix Recent trends Confidence Consequence Improving Adequate high-quality Very low Likelihood evidence and high level Insignificant Minor Moderate Major Catastrophic Deteriorating of consensus Low Almost certain Moderate Moderate High High Very high Limited evidence or Moderate Likely Low Moderate Moderate High High Stable limited consensus Possible Low Low Moderate Moderate High High Unclear Evidence and consensus Unlikely Very low Low Low Moderate Moderate too low to make an Very high Rare Very low Very low Low Low Moderate Variable assessment

112 113 Overview Dryland salinity refers to all soils in non-irrigated areas that have become saline since being cleared for agriculture. There are three basic requirements for dryland salinity to develop: a store of salt, a supply of water and a mechanism to bring both of these into contact with the ground surface (Williamson 1998). Clearing for agriculture over the last one hundred years and the replacement of perennial, deep-rooted native vegetation with the shallower rooted annual crops and pastures has increased . This recharge results in rising watertables, bringing naturally stored salts from depth to the surface. Dryland salinity occurs when the concentration of soluble salts near the surface is sufficient to reduce plant growth. In WA, dryland salinity is caused by an altered water balance, which, at some time after clearing (decades to centuries) will reach a new equilibrium and the area of land affected by salinity will cease to expand. Although the extent of salinity changes naturally over geological scales (George et al. 2008a), the process has been accelerated and enlarged by widespread clearing and land use change. Dryland salinity is a form of land degradation currently affecting both agricultural and public land in the south-west of WA. It also impacts on water resources and natural biodiversity, and can cause damage to buildings, roads and other infrastructure. More than two million hectares of broadacre farmland in Australia is estimated to be currently affected by dryland salinity (Australian Bureau Dryland salinity adversely affecting a cereal crop in the North Stirlings. of Statistics 2002), with more than half occurring in WA. The impact (Photo: Greg Lawrence, Future Farm Industries CRC). of dryland salinity on agricultural crops is variable due to differences in crop tolerance to salinity. Groundwater monitoring and modelling, coupled with regular assessments in salinity extent and risk, are required to assist in determining the various impacts and appropriate management responses.

114 115 Assessment method proportion of each hydrozone mapped as salt-affected by the Land Monitor Project in the late 1990s was analysed to assist in determining Spatial units the impacts of dryland salinity. Hydrozones were chosen as the spatial unit to assign ratings for Risk matrix and analysis dryland salinity condition and trend. Hydrozones are areas of similar climate, geology, hydrology, soils and landscapes. They are based on Dryland salinity risk was determined using an adaptation of the risk soil-landscape zone mapping units (Schoknecht et al. 2004) and reflect matrix (Table 2.7.2) developed by Spies and Woodgate (2005), which state-scaled regions with similar farming system attributes (George was based on the Australian and New Zealand Risk Management et al. 2005). Some soil-landscape zones with similar hydrological Standard AS/NZS 4360:2004 (now revised to AS/NSZ 31000:2009). characteristics were aggregated. Table 2.7.2 Salinity risk matrix

Condition in the 1990s Consequence Satellite mapping techniques are the most efficient means to assess Likelihood and map large areas of salt-affected land (Spies and Woodgate 2005). Insignificant Minor Moderate Major Catastrophic The Land Monitor Project (McFarlane et al. 2004) used satellite imagery, Almost moderate moderate high high very high with high resolution topographic data, to map areas of severely certain salt-affected land based on areas of consistently low productivity in Likely low moderate moderate high high consecutive spring satellite (Landsat TM) scenes. Possible low low moderate moderate high The project mapped severely salt-affected land over most of the dryland agricultural area of WA for two time periods: typically within Unlikely very low low low moderate moderate 1988–92 and 1996–2000. The process could detect persistent, Rare very low very low low low moderate ‘severely’ salt-affected areas with low density vegetative cover that had been degraded, but could not detect salt-affected areas that had Source: adapted from Spies and Woodgate (2005). a dense vegetative cover in spring (e.g. areas of dense barley grass or rehabilitated saltland). Hence, it underestimates salinity in high rainfall areas because much of the saline land carries permanent cover. It may also overestimate salinity in drier areas where consistently low productivity occurs for reasons other than salinity (Furby et al. 2010). Though it is now over 15 years old, the Land Monitor data is the most comprehensive, high resolution mapping of the extent of salinity currently available for the south-west of WA (Caccetta et al. 2010). The

114 115 The risk analysis was carried out by determining the likelihood and The ‘likelihood’ is the probability that salinity will have a defined impact consequence of dryland salinity expansion in each hydrozone (Figure within the specified time period. The specified time period used in this 2.7.2). The risk of dryland salinity across each hydrozone is aggregated analysis – the ‘timing of salinity’ – is the time required for hydrological and includes areas of both low and high risk. equilibrium to be reached and all of the potential dryland salinity to be

Figure 2.7.2 The process for determining salinity risk and timing. 116 117 realised. To determine the likelihood of salinity expanding and when, Table 2.7.4 Consequence categories and definitions trends in groundwater levels, climate and land management in each hydrozone were assessed (Figure 2.7.2). The likelihood categories and Consequence definitions used were those suggested by Spies and Woodgate (2005) Insignificant: Low socio economic loss, negligible impact, no and are listed in Table 2.7.3 measurable cost Minor: Small socio economic loss, little impact, low cost Table 2.7.3 Likelihood categories and definitions Moderate: Higher socio economic loss, some impact, high cost Likelihood Major: Major socio economic loss, extensive impact, major cost Rare: May occur only in exceptional circumstances Catastrophic: Enormous socio economic loss, widespread, severe Unlikely: May occur at sometime, but unlikely impact, massive cost Possible: Might possibly occur at some time The ratings for likelihood and consequence of dryland salinity in each Likely: Will probably occur in most circumstances hydrozone were analysed separately and then intersected to determine the salinity risk according to the matrix in Table 2.7.2. The salinity risk Almost certain: Is expected to occur in most circumstances ratings for each hydrozone are presented in Table 2.7.7 and graphically Source: Spies and Woodgate (2005). in Table 2.7.1, Figure 2.7.1 and Figure 2.7.8. Consequences can be considered by evaluating costs according to triple bottom line accounting standards (i.e. economic, environmental Hazard and social). These are often grouped as socio-economic impacts (Spies The Land Monitor Project used a digital elevation model (DEM) to and Woodgate 2005). The biophysical and socio-economic aspects predict the extent of salinity hazard by applying a combination of considered in this analysis were: the salinity hazard location and extent, decision tree classifications and DEM processing (Caccetta et al. 2010). the productive value of the agricultural land within the hazard area, and The salinity hazard mapped by this process is referred to as the valley potential off-site impacts on rural infrastructure, water resources and hazard (George et al. 2005) and contains four elevation classes (0.0– biodiversity (Figure 2.7.2). This methodology acknowledges that there 0.5, 0.5–1.0, 1.0–1.5 and 1.5–2.0 m). The hazard area encompasses are many uncertainties that are difficult to reduce due to the extended all areas that have a potential to develop shallow watertables, but time frames involved (Pannell 2001) and also variability within the some may not develop salinity due to a number of mitigating factors hydrozone not captured by this scale of mapping. The categories and (McFarlane et al. 2004). The mapped hazard area is best thought of as definitions of consequence are listed in Table 2.7.4. an envelope, within which the majority of future shallow watertables and/or salinity is most likely to develop, depending on soils, climate and land use. The valley hazard areas within the hydrozones were one of three factors used to determine the consequence of dryland salinity expanding in the salinity risk assessment process (Figure 2.7.2).

116 117 Groundwater trends Time series analyses of groundwater levels in 1500 bores were undertaken for the time period 2007–12. This time period was selected Groundwater trend analysis is considered one of the most efficient tools as it built upon previous analyses by DAFWA presented at the Second for predicting the extent of future dryland salinity (Coram et al. 2001) International Salinity Forum (George et al. 2008b). Bores qualified for and was used in conjunction with recent climate and land use trends to inclusion in the analyses if they were remote from any likely effects determine the likelihood of dryland salinity expanding and the timing of of salinity management treatments (drains, trees, perennial pastures) the predicted expansion. and met a minimum standard (Raper et al. 2013). Where possible, lines of best fit were derived for the period of record for all of the data; however, if there was any uncertainty, such as significant seasonal variability, trends were derived from the summer minima (Raper et al. 2013). Groundwater level trends within each hydrozone were determined and then categorised according to the dominant trend. The trend categories are illustrated in Figure 2.7.3 and Table 2.7.5 and the results are graphically presented in Figure 2.7.6.

Table 2.7.5 Categories of groundwater trend

Definition Category Summary Full F Mostly falling Groundwater levels in a majority of the bores (> 50%) in hydrozone are falling. Trend in remaining bores could be stable or rising. S Mostly stable Groundwater levels in a majority of the bores (> 50%) in hydrozone are stable. Trend in remaining bores could be falling or rising. R Mostly rising Groundwater levels in a majority of the bores (> 50%) in hydrozone are rising. Trend in remaining bores could be stable or falling. V Variable trend Groundwater levels in the hydrozone show variable trends. Bores within hydrozone have roughly equal numbers of falling, rising and stable groundwater levels.

Figure 2.7.3 Categories of dominant groundwater trend.

118 119 The trends in groundwater levels in monitoring bores within Dandaragan trends, 15% showed falling trends and 8% had stable trends between Plateau Hydrozone [6] are illustrated in Figure 2.7.4 as an example of 2007 and 2012. The hydrozone was categorised as having ‘mostly how the dominant groundwater level trends were categorised. In this rising’ trends in groundwater levels (Figure 2.7.4). hydrozone, 77% of the monitoring bores showed rising groundwater Timing of salinity Terminology and categories for timing of salinity were adapted from the Salinity Investment Framework (SIF) Phase 1 assessment of salinity impacts on agricultural land and rural infrastructure (George et al. 2005). The SIF categories were simplified and amalgamated into three categories: short, medium and long term (Table 2.7.6). Each hydrozone was assigned a timing of salinity category, which indicates how long it is expected to take the groundwater system to come to equilibrium and the area of salinity to stabilise.

Table 2.7.6 Timing of salinity terminology and categories

Time until potential salinity develops fully SIF Phase 1 terminology Simplified terminology used in this report Imminent <10 years Short term <20 years Short term 10–20 years Short–medium 20–30 years Medium term 20–75 years term Medium term 30–75 years Long term >75 years Long term >75 years Not applicable NA Not applicable NA Source: adapted from George et al. (2005).

Figure 2.7.4 Example of categorising groundwater level trend. In the Dandaragan Plateau Hydrozone, 15% of bores had falling groundwater levels, 77% had rising groundwater levels and 8% had stable groundwater levels, leading to an overall trend rating of R – mostly rising. 118 119 Current status and trends

Condition from Land Monitor analysis The area mapped as severely affected by dryland salinity, using Landsat TM was about 1.1 million hectares in 1998 and the annual rate of increase between 1988 and 1998 was about 14 000 ha (Furby et al. 2010). This area equates to just over 5% of agricultural land (McFarlane et al. 2004). Most of the salt-affected land mapped by the Land Monitor method occurs in the hydrozones occurring in the eastern parts of the south-west of WA (Figure 2.7.5). The proportion of each hydrozone mapped as areas of consistently low productivity and valley hazard areas are presented in Table 2.7.7.

Groundwater trends Groundwater levels between 2007 and 2012 were assessed in 1500 bores (Raper et al. 2013). Groundwater levels are mostly rising in the Northampton Block [2], East Binnu Sandplain [3], Dandaragan Plateau [6], South-western Zone of Ancient Drainage [11], Jerramungup Plain [22], Ravensthorpe [23], Esperance Sandplain [24] and Salmon Gums Mallee [25] hydrozones. Conversely, groundwater levels are mostly falling in the Southern Cross [9] and Western Darling Range [14] hydrozones, and mostly stable in the Eastern Darling Range [13], Southern Zone of Rejuvenated Drainage [12], Coastal Plain [15] and Warren–Denmark [18] hydrozones. DAFWA does not monitor groundwater levels in the Donnybrook– Figure 2.7.5 Areas of consistently low productivity for the south-west of Leeuwin [16] and Scott Coastal Plain [17] hydrozones and therefore WA determined by Land Monitor methodology in 1988 and groundwater trends were not assessed. Trends in groundwater levels 1998. Areas mapped as salt-affected by 1988 are shown are variable in the remaining hydrozones (1, 4, 5, 7, 8, 10, 19, 20 and 21) in orange, areas that Land Monitor determined to be salt- as there are a mixture of rising, stable and falling trends across these affected by 1998 are shown in red. The valley hazard areas zones. The dominant groundwater level trend across each hydrozone is are shown in blue. Source: Caccetta et al. 2010. summarised and listed in Table 2.7.7 and shown in Figure 2.7.6.

120 121 Timing of salinity hydrological equilibrium and the area of dryland salinity will cease to expand. More than half (14 out of 25) of the hydrozones will reach In the short term (< 20 years), the Northampton Block [2], Southern equilibrium in the medium term (20–75 years) and three hydrozones – Zone of Rejuvenated Drainage [12], Stirling Range [20], Jerramungup Dandaragan Plateau [6], Southern Cross [9] and Salmon Gums Mallee Plain [22] and most of the hydrozones west of the Darling Scarp, south [25] – will reach equilibrium in the long term (> 75 years). The results of Perth (Western Darling Range [14], Coastal Plain [15], Donnybrook– of the timing of salinity assessment is listed in Table 2.7.7 and shown Leeuwin [16] and Warren–Denmark [18]), either has or will reach graphically in Figure 2.7.7.

1 1 3 2 3 2 2007–12 Legend 4 4 Mostly Falling Short Geraldton Mostly Stable Geraldton Medium 5 5 Variable Long Mostly Rising 6 6 7 7 9 9 8 8 Merredin Merredin Perth Perth

13 13 15 15 11 14 11 14 10 10 Bunbury 12 25 Bunbury 12 25 23 23 24 16 22 24 16 21 22 21 Esperance 20 Esperance 17 20 17 18 18 19 19 0 100200 300 0 100200 300 Albany Albany Kilometres Kilometres

Figure 2.7.6 Dominant groundwater level trends (2007–12). Figure 2.7.7 Time until the hydrozones reach hydrological equilibrium and all areas of potential dryland salinity have developed. 120 121 Risk assessment The risk assessment was based on the likelihood and consequence of dryland salinity developing further in each hydrozone. The level of dryland salinity risk assessed, ranged from high, which will almost 1 3 2 Risk of expansion certainly (or is likely to) occur with moderate or major consequences, to 4 very low, which is unlikely or will rarely occur and will only have minor High consequences. The results of the risk assessment are listed in Table Geraldton Moderate 2.7.7 and shown graphically in Figure 2.7.8 and Table 2.7.1 5 Low Four hydrozones were assessed as having a high risk: East Binnu Very Low Sandplain [3], Dandaragan Plateau [6], South-western Zone of Ancient 6 Drainage [11] and the Esperance Sandplain [24]. 7 9 Nearly half of the hydrozones were assessed as having a moderate risk 8 Merredin and included Kalbarri Sandplain [1], Irwin Terrace [4], Northern Zone of Ancient Drainage [7], Northern and Southern Zones of Rejuvenated Perth Drainage [8 and 12], South-eastern Zone of Ancient Drainage [10], 13 Eastern Darling Range [13], Warren–Denmark [18], Pallinup [21], 15 11 Jerramungup Plain [22], Ravensthorpe [23] and the Salmon Gums 14 10 Mallee [25]. Bunbury 12 25 23 Just over one-quarter of the hydrozones were assessed as having a 16 22 24 low risk of dryland salinity expanding further. These were Northampton 21 17 20 Esperance Block [2], Arrowsmith [5], Southern Cross [9], Western Darling Range 18 19 [14], Coastal Plain [15], Albany Sandplain [19] and Stirling Range [20]. 0 100200 300 Only two hydrozones – Donnybrook–Leeuwin [16] and Scott Coastal Albany Kilometres [17] – were assessed as having a very low risk, though as noted, DAFWA has no surveillance bores in this area and this assessment was based on other data. Figure 2.7.8 The risk of expansion of dryland salinity.

122 123 Table 2.7.7 Summary of dryland salinity assessment

Land Monitor Trend Proportion Proportion Dominant Risk of of zone salt- of zone Hydrozone groundwater Timing of salinity Comments affected in within valley Likelihood Consequence trend salinity§ expanding‡ 1998* hazard† 2007–12‡ % % Rising, falling and stable groundwater trends are found throughout. Rising trends are observed where Kalbarri medium groundwater is less than 15 m deep, and falling and 1 0 NA variable possible moderate moderate Sandplain term stable trends where groundwater is deep (> 15 m). The salinity risk is moderate because of the extensive areas of flat plains. Prior to 2000, this hydrozone appeared to be in hydrological equilibrium with stable groundwater trends. From 2000 to 2007, drought led to significant groundwater falls. Since 2007, groundwater levels Northampton mostly short 2 NA 17 unlikely minor low have been rising but generally remain below Block rising term pre-2000 levels. The salinity risk is low because expanding salinity is unlikely and any consequences minor because of the incised topography restricting the extent of salinity. Rising groundwater trends are currently being observed. Consistent rising trends and extensive East Binnu mostly medium almost areas where the depth to groundwater is less than 3 0 NA moderate high Sandplain rising term certain 10 m make it almost certain that dryland salinity will continue to spread, with moderate consequences as there are extensive low-lying areas at risk. Rising, falling and stable groundwater trends are observed throughout. Rising trends are associated with sites located within areas of sandplain soils. Falling trends tend to be associated with heavier medium 4 Irwin Terrace 4 10 variable possible moderate moderate soil types. Stable trends are observed where term groundwater is shallow (< 2 m). The salinity risk is moderate because further salinity could develop. The consequences are moderate because of the high productive quality of the agricultural land at risk.

122 123 Land Monitor Trend Proportion Proportion Dominant Risk of of zone salt- of zone Hydrozone groundwater Timing of salinity Comments affected in within valley Likelihood Consequence trend salinity§ expanding‡ 1998* hazard† 2007–12‡ % % Rising groundwater trends observed in central and eastern parts of this hydrozone generally occur where localised perched watertables overlie the regional groundwater system. In western areas medium where extensive areas of salinity developed west of 5 Arrowsmith 1 NA variable possible minor low term Eneabba historically, mostly falling trends prevail. This zone has a low risk because further salinity could possibly develop but with minor consequence because the agricultural land at risk is low productive quality, pale deep sands. Rising groundwater trends are found throughout. At many sites the depth to groundwater is less than 10 m and salinity is actively spreading in an area south- Dandaragan mostly west of Moora (West Gillingarra). It is likely that it 6 1 NA long term likely major high Plateau rising will continue to spread with major consequences in the long term. As well as high quality agricultural land, there are high value biodiversity assets at risk of salinisation. The salinity risk is moderate because rising Northern groundwater trends continue to be observed Zone of medium throughout, therefore it is likely that salinity 7 6 30 variable likely moderate moderate Ancient term expansion will continue. The development and Drainage spread of salinity is slow and the consequence will be moderate.

124 125 Land Monitor Trend Proportion Proportion Dominant Risk of of zone salt- of zone Hydrozone groundwater Timing of salinity Comments affected in within valley Likelihood Consequence trend salinity§ expanding‡ 1998* hazard† 2007–12‡ % % The majority of bores in low landscape positions have shallow, stable watertables. There is however, a significant proportion of bores in lower landscape Northern positions still displaying rising trends. In mid to Zone of medium upper landscape positions, the majority of bores with 8 6 24 variable likely moderate moderate Rejuvenated term groundwater levels deeper than 5 m have a slightly Drainage falling trend or do not have any clear trends. Salinity is likely to expand with moderate consequences, although timing of salinity may be extended because of changes in rainfall pattern. Groundwater levels within the greenstone hills are deep (10 to 40 m). Since monitoring commenced in 2007, falling trends were evident in all lower catchment bores, with no clear groundwater trend Southern mostly 9 2 26 long term possible insignificant low observed in mid and upper catchment bores. A post- Cross falling clearing groundwater equilibrium may be decades away, although this is dependent on climate. The salinity risk is low – it could possibly occur in the future but with insignificant consequences. A large number of bores are now showing stable groundwater level trends; however, rising trends can still be observed in upland areas and in areas South- of valley hazard in the southern portion of the zone eastern Zone medium 10 6 26 variable likely moderate moderate where rainfall is highest. It is likely that salinity of Ancient term will continue to develop into the future, albeit at a Drainage reduced rate due to more dry seasons. The impact of future salinisation is expected to be moderate in the north-east of the zone, and major in the south.

124 125 Land Monitor Trend Proportion Proportion Dominant Risk of of zone salt- of zone Hydrozone groundwater Timing of salinity Comments affected in within valley Likelihood Consequence trend salinity§ expanding‡ 1998* hazard† 2007–12‡ % % Most of the currently saline land occurs on broad valley floors. An expansion of the area affected by South- salinity is likely to continue, as groundwater levels western Zone mostly medium continue to rise, particularly in areas of valley hazard. 11 9 22 likely major high of Ancient rising term A major impact is expected as there are large areas Drainage within the valley hazard not currently affected. The expected timing for the development of further salinity is 50 years or more. Most monitored catchments appear to be approaching equilibrium but there are some bores with significant rates of groundwater level rise Southern adjacent to valley hazard areas, so an expansion Zone of mostly short 12 8 24 likely moderate moderate of the area affected by salinity is likely. Localised Rejuvenated stable term impacts are expected to be moderate as fresh to Drainage brackish groundwater resources occur. The moderate salinity risk in this zone is likely to be realised within the next 20 years. Stable groundwater trends have persisted for the last 10 years and much of the zone appears to be approaching a new hydrological equilibrium. However, a significant proportion of bores in areas Eastern of valley hazard show strong rising trends, despite mostly medium almost 13 Darling 2 3 minor moderate lower than average rainfall. Furthermore, many stable term certain Range bores in or adjacent to areas already salt-affected show artesian heads. It is almost certain that the area affected by salinity will expand, though the additional area will not be great and hence the impact is assessed as minor.

126 127 Land Monitor Trend Proportion Proportion Dominant Risk of of zone salt- of zone Hydrozone groundwater Timing of salinity Comments affected in within valley Likelihood Consequence trend salinity§ expanding‡ 1998* hazard† 2007–12‡ % % Surveillance monitoring of groundwater levels is limited in this zone as it is mostly forested or has Western mostly short been reafforested. As a result, groundwater levels are 14 Darling 1 4 possible minor low falling term falling at the majority of sites. The salinity risk is low; Range expansion of salinity is possible at a local scale, with minor consequences. Groundwater is shallow over much of the plain but trends are stable, responding to seasonal rainfall. Salinisation on the coastal plain is limited to poorly drained areas on the Pinjarra Plain and coastal swales. The salinity risk is low. Expansion of the area mostly short 15 Coastal Plain NA NA possible minor low affected is possible but with minor consequences stable term depending on whether high intensity land uses move into poorly drained areas. Increasing salinity is likely in surficial aquifers. Widespread occurs in the south-west areas but is not included in this analysis. There are few areas of salinity within this hydrozone hence DAFWA does not monitor groundwater levels Donnybrook– short here. The salinity risk is very low, as the impacts 16 NA NA na unlikely insignificant very low Leeuwin term are negligible and it is unlikely to expand. Brackish groundwater seepages into water supplies occur and may be locally or seasonally relevant. The majority of the Scott Coastal Plain is forested or recently cleared and there are no large areas of salinity, hence DAFWA does not currently monitor Scott Coastal medium this zone. The hydrozone is probably close to 17 NA NA na unlikely insignificant very low Plain term hydrological equilibrium, therefore, the salinity risk is very low, due to the unlikelihood of the small area affected expanding. Salinity associated with private irrigation systems has not been assessed.

126 127 Land Monitor Trend Proportion Proportion Dominant Risk of of zone salt- of zone Hydrozone groundwater Timing of salinity Comments affected in within valley Likelihood Consequence trend salinity§ expanding‡ 1998* hazard† 2007–12‡ % % Aquifers are responsive to changes in recharge and, depending on land use and/or seasonal variability, groundwater can rise or fall accordingly. Groundwater levels have already begun to rise in Warren– mostly short response to changes in land use in some areas. 18 1 11 possible moderate moderate Denmark stable term The risk is moderate as salinity could possibly expand, due to changes in land use, with moderate consequences, particularly if streamflow salinity increases and river water quality falls below potable limits. The current low proportion of salt-affected land can be attributed to deep groundwater levels. The western section has isolated areas of salinity, many of which have been stabilised by the establishment Albany medium 19 1 24 variable possible minor low of plantations. The eastern sandplain section shows Sandplain term steady, rising groundwater trends and is cause for concern in the medium to long term. The salinity risk is low; salinity may develop further but the consequence is considered to be minor. This hydrozone already has large areas of severely salt-affected agricultural land. Over the past 15 years, many hectares of salt-affected land have been converted to saltland pastures. It appears Stirling short 20 5 24 variable unlikely minor low that the hydrozone is approaching a post-clearing Range term equilibrium; therefore, further salinity development is unlikely, with minor consequences, as large areas currently have groundwater levels at or near the surface.

128 129 Land Monitor Trend Proportion Proportion Dominant Risk of of zone salt- of zone Hydrozone groundwater Timing of salinity Comments affected in within valley Likelihood Consequence trend salinity§ expanding‡ 1998* hazard† 2007–12‡ % % The risk of salinity is moderate. The area of salinity is likely to expand but with minor consequences and will mostly be confined to drainage lines and medium hillside seeps. The impact on agricultural productivity 21 Pallinup 3 22 variable likely minor moderate term depends on where salinity develops in the landscape; it will be minor in areas with incised topography and moderate in areas of broad, flat valley floors with extensive areas of valley hazard. The risk of salinity developing is moderate, as deep groundwater levels continue to show rising trends and are likely to put further pressure on existing discharge areas. With this increased pressure, new saline areas are likely to appear, with moderate Jerramungup mostly short 22 3 15 likely moderate moderate consequences. Though the potential downstream Plain rising term impact on drainage lines that flow through the Fitzgerald River National Park has not been determined, a reduction in water quality is likely to lead to ecosystem change and loss of riparian vegetation. Most expressions of salinity are caused by hillside seeps or groundwater baseflow within waterways, which is currently occurring along most of the river channels and some tributaries. The salinity mostly medium risk is moderate. Salinity is likely to expand but the 23 Ravensthorpe <1 12 likely minor moderate rising term consequence is minor, as the area of agricultural land at risk is small. However, the off-site impacts could be considerable in relation to biodiversity, especially along the waterways in the Fitzgerald River National Park.

128 129 Land Monitor Trend Proportion Proportion Dominant Risk of of zone salt- of zone Hydrozone groundwater Timing of salinity Comments affected in within valley Likelihood Consequence trend salinity§ expanding‡ 1998* hazard† 2007–12‡ % % The area of salt-affected land has expanded since the 1990s due to a series of years of above average rainfall. Groundwater levels are either shallow and fluctuate seasonally or are rising where they are deeper (> 5 m). The salinity risk is high as salinity Esperance mostly medium almost will almost certainly expand. The consequence 24 1 21 moderate high Sandplain rising term certain is considered moderate, due to the extensive (about 20%) areas of agricultural land within the valley hazard area and the off-site impacts on the biodiversity assets, particularly those within the coastal reserves (e.g. Ramsar wetlands and national parks). Groundwater levels in the north of this zone continue to show stable trends. In the eastern part they have been falling over the last few years due Salmon mostly to below average annual rainfall. In western areas, 25 4 22 long term possible moderate moderate Gums Mallee rising where rainfall has not declined, groundwater levels continue to rise. The salinity risk is moderate and the area of salinity will possibly expand with moderate consequences in the long term. NA Not assessed * From Land Monitor areas of consistently low productivity (AOCLP 1998) † From Land Monitor average height above valley floor (AHAVF) 0–2 m, except Hydrozone 24 0-1 m and Hydrozone 25 0-0.5 m ‡ From Resource management technical report 388 (Raper et al. 2013) § From Salinity Investment Framework Phase 1 (George et al. 2005)

130 131 Discussion and implications all farmers equally and in the late 1990s, one-third (35%) of the salt- affected land mapped was managed by less than 10% (280) of farmers. Currently, more than one million hectares of agricultural and public land in the south-west of WA is severely salt-affected (McFarlane et al. Salinity also affects rural infrastructure relied upon by the agricultural 2004; Furby et al. 2010). Dryland salinity has expanded in most of this industry. About 250 km of main roads and 3850 km of local roads were region since 1998, especially following episodic rainfall events, such estimated to be affected by dryland salinity (Sparks et al. 2006). The as occurred in 1999, 2000, 2005, 2006 and 2007 (George et al. 2008b; value of rail repair and maintenance costs in 2006 was $176 million. Robertson et al. 2010). Expansion of salinity continues to take place Dryland salinity severely impacts public and private water resources despite lower than average rainfall. and biodiversity (Sparks et al. 2006). These costs are difficult to determine and therefore poorly documented. The 2000–07 dry period resulted in a reduction in the rate of groundwater level rise and a small downward trend in the depth to The implications of dryland salinity to the agricultural industry are watertables beneath some valley floors. However, since 2008, this widespread and include reduced crop yield, area of arable land, trend has largely been reversed and as a result there is no evidence of land capability and should also include the growing need to meet a reduction in the area of salt-affected land. In areas with groundwater community and market demands for environmentally responsible systems still actively filling (not near equilibrium), reduced rainfall- agriculture. recharge appears to have had little discernible impact on rising trends. Management to contain or adapt to salinity is technically feasible using Later, as these areas approach a ‘new’ hydrological equilibrium, plant-based and engineering options, though few, if any, degraded climate impacts will become the dominant controller of groundwater areas can be economically recovered. The south-west of WA will level trends and the extent of dryland salinity. The recent trends in have a significant area of salt-affected land for the long term. In the groundwater levels have been attributed to the interaction between agricultural context, salinity needs to be managed in a way which three factors: clearing, reduced rainfall, and the onset of hydrological minimises off-site impacts and enables profitable use of affected land. equilibrium (George et al. 2008a). Dryland salinity remains a potential threat to 2.8–4.5 million hectares of productive agricultural land (George et al. 2005) and depending on future climate, the area actually affected will increase. The long-term extent of salinity may take decades to centuries to develop, especially in areas where clearing was staggered, the area cleared is small (< 50%), or where watertables are deep (George et al. 2008b). The hydrozones with the highest dryland salinity risk occur mostly in the highly productive, medium to high rainfall, dryland agricultural areas. In 2009, the opportunity cost of lost agricultural production from dryland salinity in the south-west of WA was calculated to be at least $344 million per annum (Herbert, 2009). Dryland salinity does not affect

130 131 Recommendations Acknowledgements • Estimates of the current extent of salinity are undertaken using Authors: John Simons (DAFWA), Richard George (DAFWA) and Paul satellite-based systems on a decadal or similar frequency, and Raper (DAFWA) with support from members of the DAFWA water backed up by census-style surveys and farm- to catchment-scale science group, including Don Bennett, Adele Kendle, Adam Lillicrap, field-based assessments. Paul Raper, Arjen Ryder, Rosemary Smith, Russell Speed and Grant • Salinity risk assessment is based on the continued long-term Stainer. monitoring of groundwater levels and trends in the regional This chapter should be cited as: surveillance bore network. Tools, such as groundwater models Simons J, George R and Raper P (2013). ‘Dryland salinity’. In: Report and specific field analysis at key sites, are used to refine the risk card on sustainable natural resource use in agriculture, Department of assessment. Assessments are undertaken at 5 to 10 year intervals Agriculture and Food, Western Australia. at both hydrozone and catchment scales. Specific supportfor this chapter (coordination, editing, map production) • Groundwater monitoring and modelling, coupled with regular is listed in general acknowledgements. assessments of salinity extent and risk, are used by government to determine priority areas for investment and to forecast and then monitor their impact. Providing estimates of the current and Sources of information expected spatial extent and impacts of salinity at a local level also Australian Bureau of Statistics (2002). Salinity on Australian farms. allows agricultural industries and landholders to make informed Bulletin 4615.0, Australian Bureau of Statistics, Canberra, ACT. decisions on salinity management. Caccetta, D, Dunne, G, George. R & McFarlane D (2010). A • Realistic resource condition targets for salinity management are methodology to estimate the future extent of dryland salinity in the established by government, rural communities and agricultural southwest of Western Australia J. Environ. Qual. 39:26–34. businesses. The best available information on salinity risk and Coram, J, Dyson, D & Evans, R (2001). An evaluation framework for effective, economically viable management options are accessible dryland salinity. A report prepared for the National Land and Water to land managers. Government and industry regulate practices Resources Audit Dryland Salinity Project, Bureau of Rural Sciences, which exacerbate salinity impacts that fall beyond accepted Canberra. standards of practice, and provide an environment in which new industries, complementary to salinity management, can be Furby, SL, Caccetta, PA & Wallace, JF (2010). Salinity monitoring in identified and developed. Western Australia using remotely sensed and other spatial data. J. Environ. Qual. 39:16–25. George, RJ, Clarke, J & English, P (2008a). Modern and palaeogeographic trends in the salinisation of the Western Australian wheatbelt: a review. Australian Journal of Soil Research Vol 46, 751–67.

132 133 George, RJ, Kingwell, R, Hill-Tonkin, J & Nulsen, R (2005). Salinity Schoknecht, N, Tille, P & Purdie, B (2004). Soil-landscape mapping in investment framework: Agricultural land and infrastructure. Resource South-western Australia, Resource management technical report management technical report 270. Department of Agriculture, 280, Department of Agriculture, Western Australia. Western Australia. Sparks, T, George, R, Wallace, K, Pannell, D, Burnside, D & Stelfox, L George, RJ, Speed, RJ, Simons, JA, Smith, RH, Ferdowsian, R, Raper, (2006). Salinity investment framework Phase II, Salinity and land use GP & Bennett, DL (2008b). Long-term groundwater trends and their impacts series, Report no. SLUI 34, Department of Water, Perth. impact on the future extent of dryland salinity in Western Australia Spies, B and Woodgate, P (2005). Salinity mapping methods in the in a variable climate, 2nd International Salinity Forum: Salinity, water Australian context. Department of the Environment and Heritage; and and society - Global issues, local action. Adelaide, South Australia Agriculture Fisheries and Forestry. Canberra. 31 March – 3 April. Williamson, DR (1998). Land degradation processes and water quality Herbert, A (2009). Opportunity costs of land degradation hazards in the effects: waterlogging and salinisation, in Farming Action: Catchment South-west Agriculture Region - Calculating the costs of production Reaction CSIRO public. pp. 162–90. losses due to land degradation. Resource management technical report 349. Department of Agriculture and Food, Western Australia. McFarlane, DJ, George, RJ & Caccetta, PA (2004). The extent and potential area of salt-affected land in Western Australia estimated using remote sensing and Digital Terrain Models. In Proc. of Engineering Salinity Solutions, Perth, Western Australia. 9–12 November. Institution of Engineers, Barton, Australian Capital Territory. Pannell, DJ (2001). Dryland salinity: economic, scientific, social and policy dimensions. Australian Journal of Agricultural and Resource Economics 45(4). pp: 517–46. Raper, R, Speed, R, Simons, J, Kendle, A, Blake, A, Ryder, A, Smith, R, Stainer, G and Bourke, L (2013). Groundwater trend analysis in the south west land division of Western Australia, Resource management technical report 388. Department of Agriculture and Food, Western Australia. Robertson, MJ, George, RJ, O’Connor, MH, Dawes, W, Oliver, YM & Raper, GP (2010). Temporal and spatial patterns of salinity in a catchment of the central Wheatbelt of Western Australia. Australian Journal of Soil Research, Vol 48, 326–36.

132 133